RealFin: How Well Do LLMs Reason About Finance When Users Leave Things Unsaid?

2026-02-06

AI Summary

This work addresses the overconfidence of large language models (LLMs) in financial reasoning, where they often produce incorrect answers due to failure in recognizing missing critical premises. To systematically evaluate this limitation, the authors introduce REALFIN, a bilingual benchmark that generates fluent yet premise-deficient questions by deliberately removing key information from original financial exam items. A multi-task evaluation framework is proposed to assess models’ capabilities across three dimensions: answering questions, identifying missing information, and proactively abstaining from responding when necessary. This study presents the first systematic assessment of LLMs’ awareness of implicit assumption gaps in financial contexts, revealing significant reliability shortcomings in both general-purpose and finance-specific models. The findings underscore that trustworthy financial reasoning requires models to possess the metacognitive judgment to distinguish known from unknown—embodying the principle of “knowing what one knows.”

Technology Category

• Machine Learning: Large Multimodal Models (LMMs)
• Cognitive Modeling & Cognitive Systems: Conceptual Inference and Reasoning
• Natural Language Processing: (Large) Language Models

Application Category

• Economics, Online Markets and Human Computation: LLM based quality controls for crowd work
• Semantics and Knowledge: Data modeling to support human-machine intelligence, including LLMs agents, intelligent system behavior, explanations, and user-friendly interactions
• Search and Retrieval-Augmented AI: Web evaluation methodologies and metrics

Abstract

Reliable financial reasoning requires knowing not only how to answer, but also when an answer cannot be justified. In real financial practice, problems often rely on implicit assumptions that are taken for granted rather than stated explicitly, causing problems to appear solvable while lacking enough information for a definite answer. We introduce REALFIN, a bilingual benchmark that evaluates financial reasoning by systematically removing essential premises from exam-style questions while keeping them linguistically plausible. Based on this, we evaluate models under three formulations that test answering, recognizing missing information, and rejecting unjustified options, and find consistent performance drops when key conditions are absent. General-purpose models tend to over-commit and guess, while most finance-specialized models fail to clearly identify missing premises. These results highlight a critical gap in current evaluations and show that reliable financial models must know when a question should not be answered.

Problem

• financial reasoning
• implicit assumptions
• missing information
• unjustified answers
• reliable reasoning

Innovation

• financial reasoning
• missing premise detection
• LLM reliability
• REALFIN benchmark
• implicit assumptions